{"id":"W4409918354","doi":"10.1145/3733620.3733635","title":"Reservoir Sampling over Joins","year":2025,"lang":"en","type":"article","venue":"ACM SIGMOD Record","topic":"Reservoir Engineering and Simulation Methods","field":"Engineering","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Joins; Sampling (signal processing); Computer science; Programming language; Computer vision","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002478491,0.0001611273,0.0002018338,0.0002037952,0.00006540088,0.00005168027,0.0003651873,0.0001218932,0.0001350728],"category_scores_gemma":[0.0004977267,0.0001639088,0.00008640554,0.0004202178,0.00001329848,0.0001237186,0.0000970798,0.0002524032,0.00003021933],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007058772,"about_ca_system_score_gemma":0.00001725595,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002444367,"about_ca_topic_score_gemma":0.00001242907,"domain_scores_codex":[0.9991252,0.00003187025,0.0002556831,0.0001901682,0.0001214104,0.0002756629],"domain_scores_gemma":[0.9987786,0.0004049093,0.00001543387,0.0007012151,0.00003467506,0.00006512507],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000009482298,0.000008654041,0.003471225,0.0001062733,0.00005795206,0.000002465805,0.00005215707,0.9650716,0.003445818,0.0007162333,0.007571094,0.01948711],"study_design_scores_gemma":[0.0006564378,0.00002075133,0.01059894,0.0001221511,0.00002042565,0.000001160604,0.0000183407,0.6472319,0.002813505,0.007920867,0.3302418,0.0003537671],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6866974,0.0004165572,0.29855,0.0001623029,0.001388742,0.0001160724,0.000003817716,0.0007367752,0.01192833],"genre_scores_gemma":[0.8376673,0.0001864345,0.1572895,0.00009745594,0.0002749843,0.00003600971,0.00001205755,0.00006368435,0.004372542],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3226707,"threshold_uncertainty_score":0.6684007,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03661353942592897,"score_gpt":0.3200537669503436,"score_spread":0.2834402275244146,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}